Many biological molecules, especially large complexes, are highly flexible and dynamic. Although functionally important, large flexibility of many of these molecules, frequently expressed as anisotropic deformations, cannot be adequately studied using conventional means. Normal mode analysis is an effective way for describing large-scale anisotropic deformations of biomolecules, however, there exist tremendous difficulties when the size of systems of interest and the scale of deformations are getting increasingly larger. Therefore, there is a pressing need for new computational procedures to capture, with high accuracy, the large-scale anisotropic deformations for structural and functional studies. In the current funding cycle, the PI's group has successfully developed a normal-mode-based X-ray refinement protocol (NM-XREF) to improve structural refinement of biomolecules with large-scale anisotropic deformations. For the next funding cycle, the PI aims to develop a diverse array of new computational techniques for multi-scale (from coarse-grained to atomic scale) modeling of large protein deformations. The coupled use of these new methods with NM-XREF is expected to greatly enhance its efficiency in structural refinement of significant biological systems. Hypotheses: Large-scale deformations of biomolecules of functional importance are frequently anisotropic, which can be best approximated by low-frequency normal modes. Powerful new multi-scale normal mode analysis algorithms that can more accurately describe such deformations will allow more efficient structural and functional studies. Long-term Objectives: The focus has been on developing new simulation methods to represent more realistically and efficiently large-scale deformations of biomolecules in structural and functional studies. Specific Aims: 1) To develop a new modal synthesis method for mode analysis of extremely large systems. 2) To design a new algorithm for normal mode analysis in crystalline environment. 3) To unify the basis sets of normal mode analysis and Translation-Libration-Screw method. 4) To generate a new anisotropic sharpening scheme for X-ray model building. Furthermore, these new methods will be used to improve refinement of specific classes of X-ray structures that each presents its own challenges in structural study. Collectively, the implementation of a series of novel multi-scale computational methods for functional and structural study is expected to greatly benefit the entire scientific community.